Self-management interventions for chronic illness

Transcription

1 Self-management interventions for chronic illness Stanton Newman, Liz Steed, Kathleen Mulligan An increasing number of interventions have been developed for patients to better manage their chronic illnesses. They are characterised by substantial responsibility taken by patients, and are commonly referred to as selfmanagement interventions. We examine the background, content, and efficacy of such interventions for type 2 diabetes, arthritis, and asthma. Although the content and intensity of the programmes were affected by the objectives of management of the illness, the interventions differed substantially even within the three illnesses. When comparing across conditions, it is important to recognise the different objectives of the interventions and the complexity of the issues that they are attempting to tackle. For both diabetes and asthma, the objectives are concerned with the underlying control of the condition with clear strategies to achieve the desired outcome. By contrast, strategies to deal with symptoms of pain and the consequences of disability in arthritis can be more complex. The interventions that were efficacious provide some guidance as to the components needed in future programmes to achieve the best results. But to ensure that these results endure over time remains an important issue for self-management interventions. Within the next 50 years, the number of people aged 60 years and over is estimated to more than triple, and in developed countries will account for at least a third of end of the continuum, self management for chronic illnesses such as rheumatoid arthritis needs not only adherence but also behavioural change and new coping the population. 1 Associated with this rise will be an strategies, because symptoms have a great effect on increase in the rate of chronic illnesses, which are currently estimated to consume about 70% of US health-care spending; 2 it has been suggested that these illnesses will be the primary cause of death and many areas of life. Unfortunately, for many individuals, optimum selfmanagement is often difficult to achieve, as indicated by poor rates of adherence to treatment guidelines, 6 10 disability in the world by ,4 Finding the best reduced quality of life, and poor psychological wellbeing, management for chronic illnesses is therefore which are frequently reported across several chronic imperative to deal effectively with increasing numbers of patients and escalating costs. Management of most chronic illnesses is characterised by extensive responsibility that patients need to take. Whether such management entails taking medication, making lifestyle changes, or undertaking preventive action, the patients, their carers, or both make the day-today illnesses Recognition of such difficulties has led to the development of interventions that directly target aspects of patients management of chronic illness, commonly referred to as self-management interventions (SMIs). The key feature of these interventions is the aim of increasing patients involvement and control in their treatment and its effect on their lives. decisions about what actions are to be taken. Patients involvement in the management of their care is referred to as self management, which has been defined Search strategy and selection criteria by Barlow and colleagues 5 as the individual s ability to We searched MEDLINE, PsycINFO, and Embase ( ) manage the symptoms, treatment, physical and for each illness. Search terms were: ("diabetes") or ("asthma") psychosocial consequences and life style changes or ("arthritis" or "osteoarthritis" or "musculoskeletal" or inherent in living with a chronic condition. Barlow "rheum*") and ("self-management" or "self-care" or further states that for self management to be effective, it "education*" or "behav*" or "psych*" or "cognitive") and needs to encompass the ability to monitor one s ("intervention" or "program*" or "trial") and ("random*" or condition and to effect the cognitive, behavioural and "RCT"). Studies were included if (1) the study assessed an SMI: emotional responses necessary to maintain a satisfactory interventions that only provided information in a didactic quality of life. This definition implies that self format or manipulated delivery of information (eg, management is more than simple adherence to assessment of telephone consultations) were not included treatment guidelines because, in addition, it because provision of information alone has been recognised as incorporates the psychological and social management insufficient for improved self management; (2) studies were of living with a chronic illness. However, chronic published in English; (3) the study population included adults illnesses do vary in the extent to which they intrude on ( 18 years old) with type 2 diabetes, asthma, rheumatoid psychological and social worlds, and consequently in arthritis, or osteoarthritis; (4) the design was a randomised what might be necessary for effective self management. trial. The titles of all papers identified were screened. If the In some cases, such as asthma, self management can study seemed to assess an SMI, we reviewed the abstracts. We mainly consist of adherence to treatment with the examined full articles of studies that met the inclusion criteria. intention of preventing major exacerbations and consequent disruptions to quality of life. At the other Lancet 2004; 364: See Comment page 1467 Unit of Health Psychology, Centre for al and Social Sciences in Medicine, University College London, London, UK (Prof S Newman DPhil, L Steed MSc, K Mulligan MSc) Correspondence to: Prof Stanton Newman, Unit of Health Psychology, Centre for al & Social Sciences in Medicine, 2nd Floor, Wolfson Building, 48 Riding House Street, London W1W 7EY, UK [email protected] Vol 364 October 23,

2 Various disciplines have contributed to the evolution of such interventions, which in part explains their present diversity. One of the early influences was an educational approach, which principally provided patients with information in a traditional didactic format, with the expectation that greater knowledge would lead to appropriate changes in behaviour for individuals to better manage their illness. 17 However, it has become clear that although knowledge might be necessary, it is often not sufficient for behavioural change Attention turned to psychology, from which three theoretical models and their constructs have had a particular effect on the development of SMIs. In social cognitive theory, behaviour is thought to be affected by expectations, with individuals confidence in their ability to perform a given behaviour (self efficacy) particularly important. 21 This idea has been used in SMIs through teaching of skills such as problem solving and goal setting to increase self efficacy. The stress coping model 22,23 emphasises coping strategies to deal with the stress of the condition, and SMIs from this model tend to attempt to improve coping. Readiness to change is a concept from the transtheoretical model, 24,25 and refers to how prepared individuals are to make changes to their behaviour. Interventions guided by this theory focus on individuals motivation to change, and adapt their approach according to differences in participants motivation to change a behaviour. An additional influence on the development of SMIs has come from clinical psychology, particularly cognitive behavioural therapies. Central to these therapies is the importance of attempting to change how people think about their illness and themselves and how their thoughts feed into behaviour. Interventions based on this approach have particular benefit for mental health conditions, 26 and have been applied to several chronic illnesses. 15,27 30 The evolution of SMIs has accompanied the trend to move away from a system in which the health-care professional is seen as expert and the patient is a passive recipient of care, to more collaborative care in which expertise is shared between patient and professional and both parties work together to achieve best possible management. 31,32 This trend has further encouraged SMIs to change from didactic provision of information to interventions teaching problem solving and coping skills. In this review, we examine the nature and effectiveness of SMIs for chronic illness, and particularly the extent of the similarities and differences between interventions for different illnesses. We focus on three chronic illnesses: type 2 diabetes, arthritis, and asthma. These illnesses have been selected for several reasons: all three have high and increasing frequency with associated high costs to health services; SMIs and research are well developed; reviews have identified a minimum of 70 intervention studies for each illness; and it is possible to examine how the important differences in the day-to-day management and consequences of the three illnesses influence SMIs. We included only studies of adults, because in childhood and adolescence, these illnesses create additional issues for SMIs, such as parent-child relationships, child development, puberty, and education, which could constitute a further review (see for example Neinstein 42 ). Furthermore, our review is limited to randomised trials published between 1997 and 2002 to capture the content and effectiveness of fairly recent interventions. Differences between interventions We identified 21 studies for type 2 diabetes, for arthritis (15 for rheumatoid arthritis, five for osteoarthritis, and four for both types of arthritis ), and 18 for asthma. 73, SMIs for diabetes, arthritis, and asthma differed in their main objectives. The objectives of the interventions for asthma tended to focus on prevention of acute exacerbations through recognition and avoidance of the triggers that provoke asthma, monitoring of symptoms and consequent adjustment of medication, and improving adherence to medication; very few addressed emotional aspects of coping with asthma. For diabetes, the objectives of the interventions tended to be more diverse, with most focusing on lifestyle issues and others on management of stress. The aim of interventions for arthritis was also quite broad with the main focus of reducing pain and improving physical and psychological functioning. These objectives affect a range of methodological and content issues. Methodological issues Most studies examined whether one or more SMIs led to better outcomes than standard care or basic information (15 [71%] for diabetes, table 1, 22 [92%] for arthritis, table 2, and 16 [89%] for asthma, table 3). The rest directly compared two or more SMIs, which could be instructive for measuring whether different components differ in their effectiveness. For all illnesses, the number of participants recruited varied widely ( for diabetes, for asthma, for arthritis). Small samples are unlikely to produce reliable results because of insufficient power. This concern seems to have been recognised in the diabetes studies, which had a mean of 65 participants per intervention group in those completed in 2002, compared with 19 before However, similar patterns were not apparent for asthma or arthritis. In arthritis, particularly large samples were used in four studies that recruited volunteers from the community or from primary care, suggesting that method of recruitment could be an important determinant of sample size. Comparability of different types of SMIs is likely to be affected by the greater power of larger studies to detect significant effects compared with smaller studies Vol 364 October 23, 2004

3 The time demands of SMIs could result in low participation and high rates of attrition. Analysis of participation rates is difficult because many studies fail to report them, and if recruitment includes some form of advertisement or open invitation, calculation of participation rates is not possible. However, knowledge Author, year n (completing), Recruitment Follow-up Groups (intervention group as Duration Individual Delivered by Outcomes (country) mean age, % male, described by study authors)* or group assessed other characteristics Agurs-Collins, (55), 61 7years, Hospital clinics, 3 and 6 months (1) Standard care 1 session Group Dietician and Clinical assessment 1997 (USA)33% male, all community adverts from baseline (2)Weight loss and exercise 1 5 h per week for 12 weeks and Both exercise physiologist behaviour >55 years, 1 5 h per fortnight for 3 months African-American and 1 individual session Smith, (16), 62 years, Advert, letter from Post (1) al weight control 16 sessions over 4 months Group Multidisciplinary team Clinical assessment (USA)0% male, all physicians intervention (2)al weight control and 19 sessions over 4 months Both Multidisciplinary team >50 years, obese motivational interviewing Samaras, (not stated), Outpatient clinic 6 and 12 (1) Standard care >1 h per month for 6 months Clinical assessment (Australia)61 years, 38% male, months from (2)Exercise and support group Group Multidisciplinary none exercising baseline team Jablon, (20), 59 years, Outpatient clinics Post (1) Standard care 8 1-h sessions over 4 weeks and Clinical assessment (USA) 50% male intervention (2) Progressive relaxation and practice with 20-min tape twice Individual Not stated Psychological biofeedback per day wellbeing Henry, (19), 60 years, Outpatient Post (1) Standard care 1 5 h per week for 6 weeks Clinical assessment (Australia)47% male clinics intervention (2)Cognitive behavioural stress Group Psychologist Psychological wellbeing management Aikens, (not stated), Urban diabetes 2 months (1) Standard care Clinical assessment (USA) 61 0 years, clinic (2) Relaxation training 6 1-h sessions over 8 weeks and Group Psychologist Psychological wellbeing 39% male practice with 30-min tape Vazquez, (not stated), Community, 3 months (1) Standard care 12 weekly sessions and 8 Group Nutritionist and (USA) Caribbean Latinos, hospital, adverts, (2) Nutrition intervention bimonthly sessions psychologist obese participant referral Lustman, (42), 55 years, Advertised to Post (1) Diabetes education 2 h per week for 10 weeks Individual Educator Clinical assessment 1998 (USA) 88% male, all physicians and in intervention, (2) Diabetes education 3 h per week for 10 weeks Individual Educator Psychological wellbeing depressed media 6 months and cognitive behaviour therapy and psychologist Ridgeway, (38), 64 years, Physicians 6 and 12 (1) Standard care Clinical assessment 1999 (USA) 29% male recommended to months (2) Education/behaviour 1 5 h per month for 6 months Both Nurse and dietician patients modification programme and 1 session at 12 months Kirk, (23), 49 years, Database and 5 weeks from (1) Standard information 30-min session Individual Researcher (UK)44% male, selected outpatients baseline (2)Exercise consultation 30-min session Individual Researcher Psychological wellbeing by readiness to change McKay, (68), 52 years, and online 8 weeks from (1) Internet information only Varied Individual Internet (USA) 47% male invitations baseline (2) Internet exercise programme 8 weeks Individual Internet and Psychological wellbeing occupational therapist Miller, (92), 73 years, Adverts, news Post (1) Standard care Clinical assessment (USA) 47% male, all >65 years letters intervention (2) Nutrition education programme h per week for 10 weeks Group Dietician McKay, (133), 59 years, Letters from 3 month - (1) Internet information only Varied over 3 months Individual Not applicable Clinical assessment 2002 (USA) 47% male primary care (2) Internet self-management Contacts twice per week for Individual Professional with physician 3 months dietary experience Psychological wellbeing (3) Internet peer support Varied over 3 months Individual Not applicable (4) Internet self-management and Varied over 3 months Individual Professional with peer support dietary experience Levetan, (128), 59 years, Education 6 months (1) Standard care Clinical assessment 2002 (USA) 33% male, diabetes programme from baseline (2) Computerised goals Individual Educator type not stated (we assume type 2) 10-min phone call Kenardy, (not stated), Clinic Post (1) Non-prescriptive therapy 1 5 h per week for 10 weeks Group Psychologist 2002 (Australia)55 years, 0% male, intervention (2)Cognitive behaviour therapy 1 5 h per week for 10 weeks Group Psychologist Clinical assessment all binge eaters Psychological wellbeing Keyserling, (varied), 59 years, Clinician invitation 6 and 12 (1) Minimum intervention Clinical assessment 2002 (USA) 0% male, African- months (2) Clinic intervention Monthly sessions for 4 months Individual Nutritionist American women (3) Clinic and community Intervention Monthly sessions for 4 months Both Nutritionist and peer Psychological wellbeing >40 years and 2 group sessions and counsellor telephone calls (continues next page) of participation is important since it indicates the extent to which results can be generalised. 110 Attrition rates varied widely, ranging from 0% to roughly 50% for each illness (see for example Jablon and others 46 vs Rickheim and others 63 for diabetes, Berg and others 88 vs Blixen and others 101 for asthma, and Hopmanwww.thelancet.com Vol 364 October 23,

4 (table 1, continued) Author, year n (completing), Recruitment Follow-up Groups (intervention group as Duration Individual Delivered by Outcomes (country) mean age, % male, described by study authors)* or group assessed other characteristics Glasgow, (285), 58 years, Letter from primary- 12 months (1) Brief dietary intervention 1 2 h at baseline, 3 and 6 months Individual Nurse, dietician, Clinical assessment 2002 (USA) 43% male care physician educator, or psychologist (2) Brief dietary intervention and 1 2 h at baseline, 3 and Individual community resources 6 months, and 7, min phone calls over 12 months (3) Brief dietary intervention and 1 2 h at baseline, 3 and Individual telephone follow-up 6 months, and 7, min phone calls over 12 months (4) Brief dietary intervention, 1 2 h at baseline, 3 and Individual telephone follow-up, and community 6 months resources Brown, (not stated), Databases of other 6 and 24 (1) Standard care 2 h per week for 3 months, Group Nurse, dietician, and Clinical assessment 2002 (USA) 54 years, 19 5% male, research studies months from (2) Self-management education 2 h per fortnight for 6 months, community workers Mexican American baseline and 2 h per month for 3 months Trento, (90), Clinic database 24, 36, and (1) Standard care 15 sessions over 4 years Group Physician and Clinical assessment 2002 (Italy) 62 years, 48 months (2) Interactive group visits educator 54% male Surwit, (72), 57 years, Adverts, medical 2, 4, 6, and (1) Diabetes education 0 5 h per week for 5 weeks Group Not stated Clinical assessment 2002 (USA) 58% male facilities, education, 12 months from (2) Stress management and diabetes Group Not stated and support groups baseline education 0 5 hrs per week for Psychological wellbeing 5 weeks Rickheim, (92), 52 years, Referred by 6 months from (1) Individual diabetes education 2 h at baseline, 1 h at 2 weeks, Individual Nurse and nutrition Clinical assessment 2002 (USA)66% male primary care baseline and at 3 and 6 months specialist (2) Group diabetes education 3 h at baseline, 2 h at 2 weeks, Group Nurse and Psychological wellbeing and 1 h at 3 and 6 months nutrition specialist Clinical assessments included HbA1c as indicator of glycaemic control. Other clinical assessments (eg, blood lipids, weight, body-mass index) were excluded. included measures of diabetes self-management behaviours (eg, diet, exercise, home blood glucose monitoring), but behaviours specifically related to intervention (eg, use of community resources) excluded. Psychological wellbeing included any composite measures of wellbeing or assessments of mood (eg, depression or anxiety). Perceived stress also included in this category. Both diabetes specific and generic quality of life measures included under quality-of-life category. One measure of illness intrusiveness included under this subsection to reflect reporting in original paper. Studies with mixed populations in which results were not explained separately were excluded (includes Glasgows studies). Description of every intervention aims to provide reflective summary of what each entailed; all elements of all interventions could not be included. Where duration of session is range, higher figure taken. Where mean age is presented by group, mean of those presented taken, same applies for sex. n=number randomised (available at follow-up). Post intervention=follow-up period from post intervention. *Content varies significantly despite similar titles (see webtable 1 at Full details in webtable 1. Table 1: Type 2 diabetes Rock and Westhoff 83 vs Cronan and others 79 for arthritis). Although long follow-up has been associated with high attrition in some instances, 79 this was not always the case. 111 The time patients spend on the intervention might be expected to change attrition, but little association was seen for any illness. For example, some SMIs for diabetes with the greatest intensity did have high attrition rates, 50,51 but others with short durations had even higher rates. 62 There was, however, some indication that interventions spread over long periods had high attrition rates. 44,63,75,106 The factors leading to attrition are clearly complex and worthy of investigation to improve the benefits of SMIs, but characteristics of individuals most likely to drop out should also be investigated to ensure that these interventions are targeted most effectively. Most (67%) SMIs for arthritis were delivered in a group setting, whereas for diabetes and asthma, individual and group settings were used in roughly similar proportions. The arguments for using group intervention include reduced costs and the potential value of group learning. Individual interventions are often justified on the basis that the intervention can be tailored to individuals needs, and they might also be easier to integrate into clinical practice. Evidence on the effectiveness of group or individual delivery is scarce because comparisons across studies are confounded by many other differences. One study in diabetes 63 specifically compared group and individual SMIs and reported that the group-based intervention resulted in greater improvements in blood glucose at 6 months follow-up; however, no differences were recorded for any other outcomes. Almost all programmes were delivered face to face, although the telephone was also used in some. One intervention for diabetes 56 used telephone alone, and one for asthma 107 and one for arthritis 84 were self administered with a workbook and an audiotape. All three showed some evidence of effectiveness, suggesting that, with some groups, a more remote approach could yield beneficial outcomes. The internet, which is likely to become increasingly used as a costeffective medium through which self management can be delivered, was used by one group, but no improvements were recorded in any of the outcomes measured. 53,55 Although it is too early to form conclusions, remote approaches could be particularly appealing for certain groups, for example, those in fulltime employment Vol 364 October 23, 2004

5 Author, year n (completing), Recruitment Follow-up Groups (intervention Duration Individual Delivered by Outcomes mean age, % male, group as described by or group assessed other characteristics study authors)* Rheumatoid arthritis Lindroth, (96), 55 years, Referral by 3 months, (1) Standard care 2 5 h per week for Group Multidisciplinary team of health Symptoms 1997 (Sweden) 12% male rheumatologists 12 months (2) Problem-based 8 weeks professionals doctor, nurse, Functioning education physiotherapist, occupational therapist, social worker, and dietician. Bell, (127), 56 years, Patients referred After 12 (1) Standard care At least 3 h or Individual Physical therapists Clinical assessment (Canada)20% male for physical weeks from (2)Community based visits 4 therapist over Symptoms therapy baseline physical therapy and 6-week period. education (information booklets, physical therapy, and goal-setting) Brus, (55), 59 years, Rheumatology 3 months, (1) Standard care Four 2-h meetings Group Unclear Clinical assessment (Netherlands)20 % male, rheumatoid outpatients 6 months, (2)Education programme during first month Symptoms arthritis for <3 years 12 months plus reinforcement Functioning meetings after 4 and 8 months Psychological wellbeing Hammond, (UK) 35 (33), 55 years, Rheumatology 12 weeks (1) Standard care 2 h per week for Group Rheumatology occupational Clinical assessment (NB: crossover trial; 17% male outpatients from (2) Educational-behavioural 4 weeks plus an therapist Symptoms only first, controlled baseline joint protection programme optional home visit Functioning phase described here) Helliwell, (77), 56 years, Rheumatology Post (1) Standard care 2 h per week for Group Non-medical health Clinical assessment 1999 (UK) 34% male, rheumatoid outpatients intervention, (2) Education programme 4 weeks professionals Symptoms arthritis for <5 years 12 months Functioning post baseline Leibing, (55), 53 years, Rheumatology Post (1) Standard care 1 5 h per week for Group Psychotherapists Clinical assessment 1999 (Germany)22% male (NB: results outpatients intervention (2)Cognitive behavioural 12 weeks Symptoms reported on only 39 treatment Functioning patients for whom Psychological wellbeing medication changes were matched) Lundgren, (60), 57 years Patients Post (1) Standard care 30 min twice a week Group Taped instructions. Physical Symptoms (Sweden) 1999 (median), 23% male registered at intervention, (2) Relaxation training for 10 weeks therapist present to assist Functioning rehabilitation 6 months, (5 weeks muscle centre 12 months relaxation followed rheumatology by 5 weeks pain unit reduction techniques) Scholten, (64), 48 years, Consecutive Post (1) Standard care 9 afternoons over Group Multidisciplinary team: Functioning 1999 (Austria)26% male patients intervention, (2)Arthritis training two weeks plus rheumatoligists, (NB: crossover 6 weeks, programme monthly meetings orthopaedists, physiotherapists, Psychological wellbeing trial; only first, 12 months psychologists, social workers controlled phase from described here)baseline Smyth, (49), 51 1 years Volunteers, 4 months (1) Writing about emo- 20 min on 3 Individual Not applicable Clinical assessment 1999 (USA)(median), 29% male adverts in local tionally neutral topics consecutive days papers, hospitals, (2) Writing about 20 min on Individual Not applicable and medical stressful event 3 consecutive days practices Hammond, (123), 51 years Referral from 6 months, (1) Standard education 4 sessions of 2 h each Group (1) Nursing, medical, Clinical assessment 2001 (UK) (median), 24% male rheumatology 12 months control group occupational therapy Symptoms outpatients and physiotherapy staff Functioning (2) Educational-behavioural 4 sessions of 2 h each Group (2) Rheumatology joint protection programme occupational therapist Hill (UK) 100 (63), 63 years Rheumatology Post (1) Standard care Both 30 min per Individual Rheumatology nurse Clinical assessment (median) in outpatients intervention (2) Education programme month over practitioner experimental group 6 months, total Symptoms and 62 years (median)7 visits in control group, 27% male Multon, (128), 58 years, Veterans hosp- Post (1) Standard care control Symptoms (USA) 56% male ital, university intervention (2) Attention control Both interventions Individual (2) and (3) both medical centre, 3 months, (3) Stress management lasted 1 5 h per week delivered by counsellors private rheumat- 15 months maintenance session for 10 weeks plus a with masters degrees ology practice at least every 3 months in psychology plus for 15 months computerised multimedia component (continues next page) Vol 364 October 23,

6 (table 2, continued) Author, year n (completing), Recruitment Follow-up Groups (intervention Duration Individual Delivered by Outcomes mean age, % male, group as described by or group assessed other characteristics study authors)* Sharpe, (45), 55 years, Rheumatology Post (1) Standard care 1h per week for Individual Psychologist Clinical assessment (UK)30% male, <2 years outpatients, 6 months (2)Cognitive behavioural 8 weeks Symptoms disease history clinics intervention Functioning Psychological wellbeing Evers, (59), 54 years, Rheumatology Post (1) Standard care 10 bi-weekly 1 h Individual (1) Rheumatology consultant Clinical assessment (Netherlands) 28% male, rheumatoid outpatients intervention (2) Cognitive behavioural sessions plus 1 booster (2) Rheumatology consultant Symptoms arthritis for <8 years 6 months therapy session 4 weeks later and therapists trained in Functioning treatment module Psychological wellbeing Freeman, (53), 51 years, 15% Rheumatology 3 months, (1) Standard arthritis Both 2 h per week Group Both by multidisciplinary Clinical assessment (UK)male, newly diagnosed clinics 6 months education control for four weeks team Symptoms (2) Cognitive behavioural Functioning arthritis education Psychological wellbeing Osteoarthritis Cronan, (178), 70 years, Written 1, 2, and (1) Standard care 2 h per week for Group Education delivered by Symptoms (USA) 36% male invitation to 3 years (2) Education 10 weeks, followed by professional health educators. members of post baseline (3) Social support 2 h per month for No staff present at social health (4) Education and social 10 months support sessions maintenance support organisation Keefe, (82), 63 years, Volunteers Post (1) Education with spousal All 2 h per week Group Psychologist and nurse Symptoms (USA)39% male intervention support control for 10 weeks Functioning (NB: follow-up of 6 months (2) Coping skills training Psychological wellbeing Keefe et al 1996) 12 months (3) Spouse assisted coping skills training Lord, (UK) 170 (126), 63 years, Primary care 1, 3, 6, and (1) Standard care 1 h per week for Group Research nurse 12-month results: 27% male 12 months (2) Education programme 4 weeks Symptoms post baseline Functioning Psychological wellbeing Maurer, (98), 65 years Outpatient 8 and (1) Exercise 3 times per week for Group Symptoms (USA) (median), 58% male clinic 12 weeks (2) Education 8 weeks (2) Rheumatologist, dietician, Functioning from 4 sessions social worker, psychologist baseline Hopman-Rock, (105) with Newspaper and Post (1) Standard care 2 h per week for Group Peer educator, physical therapist, Symptoms 2000 (Netherlands) confirmed diagnosis of television intervention (2) Health education and 6 weeks general practitioner (GP), Functioning osteoarthritis, 65 years, adverts 6 months exercise programme occupational therapist 17% male Rheumatoid arthritis and osteoarthritis combined Fries, (USA)1099 (809), 64 years, Health 6 months (1)Standard care Sent 3-monthly for Individual Self-instruction Clinical assessment 28% male maintenance from baseline (2) Mail-delivered individually 6 months Symptoms organisation, tailored programme Functioning physician referral, and participants in a general health education programme Lorig, (USA)331 (285), 62 5 years, Not reported 4 months (1)Standard care 2 h per week for Group Trained lay leaders Symptoms (NB: only randomised 16% male, Spanish post (2) Spanish-language 6 weeks Functioning part of study reported speakers baseline adaptation of the arthritis here)self-management Psychological wellbeing programme Barlow, (423), 58 years, Arthritis 4 months (1) Standard care 2 h per week for Group Trained lay leaders Symptoms (UK)16% male charity branch post baseline (2)UK evaluation of the 6 weeks Functioning networks, arthritis self-management information programme placed in GP surgeries, outpatient clinics, and media announcements Solomon, (113), 65 years, Primary care 4 months (1) Control group 2 h per week for Group Trained facilitator Symptoms (USA) 29% male post (2) Arthritis self- 6 weeks Functioning baseline management programme *Content varies significantly despite similar titles (see webtable 1 at Full details in webtable 2. Table 2: Arthritis Vol 364 October 23, 2004

7 Author, year n (completing), mean Recruitment Follow-up Groups (intervention Duration of Individual Delivered by Outcomes (country) age, % male, other group as described by intervention or group assessed characteristics study authors)* Berg, (54), not reported, Brochures placed in pharmacies Post- (1) Standard care 2 h per week for Group Nurses Clinical assessment 1997 (USA) 36% male and physician offices; radio intervention (2) Self-management 6 weeks Symptoms and newspaper announcements Kauppinen, 89,90, (134), 44 years, Outpatient clinic 1, 3, and (1) Basic education 1 session Both Respiratory nurses Clinical assessment 1998, 1999, % male, newly 5 years from (2) Intensive patient 30 min individual every third Respiratory nurses, (Finland) diagnosed patients baseline education month for 1 year plus a 2-h group chest physician, and session between the 6 and 9 physiotherapist month visits Turner, (92), 34 years, Primary care 6 months (1) Symptom based 30 min per month for Individual Nurse Clinical assessment 1998 (Canada) 47% male database adverts from baseline self-management 6 months Individual Symptoms (2) PEF based Nurse self-management 0 5 h per month for 6 months Bailey, (221), 48% aged Pulmonary and 6, 12, 18, and (1) Standard care Individual Both Health educator (USA)>40, 31% male critical care clinics 24 months (2)Core-elements (2)One min session plus Symptoms from baseline programme one telephone call 1 week later Functioning plus follow-up letter 2 weeks later (3) Self-management (3) 1 h individual session plus 2 group sessions (duration not specified) followed by 2 telephone calls plus 3 follow-up letters over a 6 8 week period George, (50), 29 years, Patients admitted from 6 months (1) Standard care "Repetitive sessions" during Not Asthma nurse (USA)21% male emergency department with (2)Inpatient hospital stay, telephone call 24 h clear specialist acute exacerbation of asthma education programme post discharge, reinforcement session 1 week post discharge de Oliveira, (42), 40 years, Outpatients database 6 months (1) Standard care 1 visit per month for 6 months Both Physician Clinical assessment (Brazil) 12% male, patients with from baseline (2) Education group and 1 h at months 3 and 4 Symptoms moderate to severe asthma Smyth, (58), 41 years, Volunteers, adverts in local 4 months (1) Writing about emo- 20 min on 3 consecutive days Individual Not applicable Clinical assessment (USA) 26% male papers, hospitals, and medical tionally neutral topics practices (2) Writing about 20 min on 3 consecutive days Individual Not applicable stressful event Cote, (149), 37 years, Following admission or clinic 12 months (1) Control group Symptoms 2000 (Canada)36% male, patients with visit to tertiary care unit (2)Education and 1 h Individual Educator moderate asthma symptom-based requiring daily monitoring corticosteroids (3) Education and peak 1 h Educator flow monitoring Gallefoss, 97,98,99 78 (71), 43 years, Patients with mild to moderate 12 months (1) Standard care Four 2-h group sessions on 2 Both Doctor, pharmacist, Clinical assessment , % male, (this study asthma, recruited at outpatient (2) Patient education separate days plus two to four nurse, (Norway) also included patients chest clinic and self-management 40 min individual sessions physiotherapist with chronic obstructive pulmonary disease, but those results not reported here) Levy, (79 87%), 42 years, Patients attending emergency 3 and 6 (1) Standard care 1 h then two 0 5 h sessions at Individual Nurse Clinical assessment 2000 (UK) 38% male department or admitted to months from (2) Patient education 6-week intervals hospital baseline Blixen, , 36 years, 29% male, Patients admitted for asthma 3 and 6 (1) Standard care Three 1-h sessions in hospital Individual Nurse (USA)African-Americans months post (2)Education Psychological admitted to hospital discharge wellbeing Cote, (98), 34 years, Patients visiting emergency 2 weeks, (1) Standard care Clinical assessment (Canada) 41% male room or outpatient 6 months, and (2) Limited education (2) Not specified Individual On-call physician clinic for acute exacerbation 12 months (3) Structured (3) One session within 2 weeks Individual Not specified of asthma from baseline education of randomisation plus one or group reinforcement session at 6 months Klein, , 245 (174), 44 years, Outpatient database 4, 8, 12, 18, (1) Self-management 1 5 h per week for 3 weeks Group Nurse Clinical assessment van Palen, % male, patients with and 24 education Symptoms (Netherlands)stable, moderate to months from (2)Self-management Group Nurse severe asthma baseline education with self- treatment guidelines 1 5 h per week for 3 weeks (continues next page) Vol 364 October 23,

8 (table 3, continued) Author, year n (completing), mean Recruitment Follow-up Groups (intervention Duration of Individual Delivered by Outcomes (country) age, % male, other group as described intervention or group assessed characteristics by authors)* Schmaling (25), 39 years, Referred by physician because 1 week (1) Education 1 brief session Not clear Clinical assessment (USA)48% male of difficulties taking medication (2) Education plus As for group 1 plus an additional Individual Masters level motivational min session of therapists interviewing motivational interviewing Couturaud, (54), 38 years, 32% Outpatient clinics Post (1) Standard care h at baseline and at Individual Nurse Clinical assessment (France) male, patients with intervention (2) Education and months 1, 3, 6, 9, and 12) Symptoms moderate to severe (ie, 12 months self-management asthma after baseline) Hockemeyer, (54), 21 years, 46% Psychology classes and Post (1) Placebo control Workbook to be used over Individual Both Clinical assessment 2002 (USA) male, university students campus community intervention (2) Stress management 4 weeks (approx 1 h per week) Individual self-administered Psychological programme Workbook to be used over wellbeing 4 weeks (approx 1 2 h per week) Marabini, (not stated), 51 Outpatient clinic 3 months from (1) Standard care Three 2-h Group Physician Clinical assessment (Italy) years, 47% male, patients baseline (2) Patient education sessions Symptoms with persistent mild, moderate, or severe asthma Pernerger, (115), no mean Emergency department or 6 months (1) Standard care 1 25 h per week for 3 weeks Group Physician and Clinical assessment 2002 (Switzerland) age, 60% male hospital wards from baseline (2) Patient physiotherapist education FEV=forced expiratory volume. PEF=peak expiratory flow. *Content varies significantly despite similar titles (see webtable 3 at Full details in webtable 3. Table 3: Asthma In all but three studies, which were delivered by lay leaders, health-care professionals led the interventions. Lay leaders have the benefit of acting as role models and being less costly, but health-care professionals are more able to address factual issues related to an illness. Presently, little evidence suggests which approach is more effective. Two studies that made direct comparisons between lay-led and professional-led programmes for arthritis 112,113 showed no improvement in pain or disability with either approach but showed differential changes in other outcomes. There is also insufficient evidence to show whether certain groups or professions are better placed to deliver SMIs. An important determinant of effectiveness is likely to be training, particularly when complex skills such as cognitive behavioural techniques are used, but information about training of course leaders was not commonly reported. Although the duration of SMIs varied for all illnesses, asthma interventions tended to be far briefer (never more than 12 h), by comparison with both diabetes (maximum 58 h 60 ) or arthritis (maximum 40 h 79 ). This difference probably results from the different objectives of care between the illnesses, with asthma tending to focus specifically on monitoring of symptoms and better adherence to medication to decrease hospital admissions, whereas both diabetes and arthritis focused on various behavioural changes including both lifestyle and cognitive components. Content of intervention The theoretical approach on which SMIs were based was more often mentioned in studies of diabetes and arthritis than of asthma, for which programmes were rarely driven by theory but tended to be principally information-based and instructional, with only a few studies incorporating techniques to address barriers to effective self-management. Diabetes and arthritis studies varied in the extent to which they explicitly stated the theory upon which the SMI was based. Sometimes only components of a theory were mentioned (eg, selfefficacy) without definition of the underlying theory. When indicated however, the interventions fell into three main areas social learning, a cognitive behavioural model, or an educational model expanded to incorporate other components such as social support, exercise, or practical tasks. Social learning theory was commonly applied and led to use of problem-solving and goal-setting to enhance participants self-efficacy and to increase health-related behaviours such as exercise, diet, blood glucose monitoring in diabetes, and exercise, joint protection, and techniques for cognitive pain management in arthritis. Other studies focused more specifically on a cognitive behavioural approach to target pain and physical function and to improve coping in arthritis or, for example, management of eating behaviours 58 or depression 50 in diabetes. Clearly, however, there is some overlap between these approaches and some studies used techniques from more than one approach, making assessment of the benefit of any one theory complex. The use of theory in self-management needs to become more explicit if we are to establish which theoretical approach is most valuable for which illnesses (see also Norris and others 39 ). This is not to say that a Vol 364 October 23, 2004

9 single theoretical approach will suffice in any area, and a more pragmatic approach could be more likely to yield results, but it remains necessary to be clear and consistent about which theoretical approaches and concepts have been used. Doing so could be aided by an electronic repository in which a full description of the SMI can be recorded, since journal editors often do not allow more detailed descriptions of the interventions because of space limitations. Comparisons of different SMIs allows identification of components or approaches that might be more effective than others. For diabetes, three of six studies that compared different SMIs found some differences between them. One showed that a more intensive intervention led to greater improvements in exercise than a less intensive one, 58 a second reported that a group intervention improved blood glucose control compared with individual intervention, 63 and a third showed that adding motivational interviewing to a behaviour-modification programme led to improvement in HbA 1c (glycosylated haemoglobin) and increased monitoring of blood glucose. 44 For osteoarthritis, comparison between three groups observed some changes favouring the inclusion of training in coping skills. 80 In another study with osteoarthritis, greater reductions in pain were reported in an exercise group than in an education group after intervention, but findings were inconsistent for other outcomes. 82 For asthma, no differences were seen with or without action plans, 103 and little difference was recorded between use of symptoms or peak expiratory flow to guide use of medication. Although comparison of SMIs has great potential for identifying active components, the findings discussed show that these studies are few and do not yet give a coherent picture in any of the three illnesses. For diabetes, comparison between different SMIs has tended to increase over time, with five of the six studies comparing interventions in This trend might indicate a general acceptance of SMIs in diabetes research, and that the research agenda has moved to increasing their efficacy by establishing the most effective components. Outcomes To examine whether particular outcomes were focused upon, those in each study were classified into seven broad categories (table 4). Whereas the most commonly assessed outcome for diabetes and asthma was clinical assessments, the outcome measure most frequently assessed in arthritis was self-reported symptoms. Most studies (55 [87%]) assessed more than one outcome. Many studies included outcomes that they had not specifically targeted in their programme, and this practice could dilute their overall effectiveness. To accurately assess an SMI, it is important to link the outcomes measured to those targeted for change. For diabetes, 11 (61%) of 18 of the studies that measured HbA 1c showed some evidence of effectiveness at some point, as has been reported in previous reviews. 39,114 Although concern was expressed in these reviews about the long-term retention of these effects, the studies reviewed here indicated that in four of the seven studies that assessed HbA 1c beyond 6 months, 60 62,111 sustained improvements were measured. All three studies showing long-term improvements were group-based, and although two used problem-solving amongst other components, the third used an approach that emphasised stress management. Superficially, this pattern suggests that there are different approaches to design of effective SMIs that have long-term effectiveness in type 2 diabetes, although the level of description in the reports could mask their similarity. The most commonly reported clinical outcome in arthritis was physician s assessment of the number of painful and swollen joints. Three of the 12 studies (25%) that assessed some form of joint count had significant findings. 73,76,84 The techniques in these studies were diverse: one used cognitive behavioural therapy, 76 another expressive writing, 73 and a third 84 a programme delivered by mail. A recent Cochrane review 115 reported that although SMIs for adults with asthma had little effect on lung function overall, this outcome was better in those who adjusted their medication using a written plan than in those whose medication was adjusted by a doctor. 115 In our review, 57% (eight of 14) of studies showed some improvement in lung function. Although most (five of eight) used a combination of education with an action plan, 91,92,97,100,103 others that used this approach did not find any improvements in lung function. 106,108 Additionally, a writing intervention for emotional expression 73 and a stress management intervention 107 also improved lung function, suggesting that methods directed at stress and emotions as well as behaviour can be successful in improving lung function in asthma. Symptoms and functioning were common outcomes in both arthritis and asthma but were not measured at all in diabetes, because of the nature of the illness. Overall, roughly 40% of SMIs for arthritis showed some improvement in self-reported symptoms, as did a similar proportion for measures of disability. Of the three arthritis groupings, the least favourable outcomes were in rheumatoid arthritis, for which few studies over the 5-year period showed an effect on pain or disability. Previous reviews of the same outcomes in such arthritis have generally reported that SMIs have a small but significant short-term effect, although the changes tend not to be maintained in the long term. 19, In these studies, some evidence suggested improvements in pain beyond 6 months in one study 74 and for disability in two. 72,74 In agreement with a previous review, 116 a greater effect on pain was identified for osteoarthritis, with four of five studies reporting some benefit. The reasons for Vol 364 October 23,

10 Arthritis (n=24) Asthma (n=18) Diabetes (n=21) Clinical and laboratory assessments* 12 (50%)14 (78%)18 (86%) Self-reported symptoms 22 (92%)8 (44%)0 Self-reported functioning 18 (75%)1 (6%) 0 Psychological wellbeing 10 (42%)2 (11%)10 (48%) 7 (29%)12 (67%) 7 (33%) 12 (50%)14 (78%)13 (62%) Use of health care** 6 (25%)11 (61%) 0 *Includes for arthritis some form of joint count either alone or in combination; includes for diabetes a laboratory measure of glycosylated haemoglobin, normally hba1c; and includes for asthma an objective measure of lung function (eg, peak expiratory flow, forced expiratory volume, or forced vital capacity). Includes all self-reports of symptoms (eg, pain, breathing difficulties). Includes self-reported measures of disability (eg, Health Assessment Questionnaire) and assessments of effect of activities of daily living on work. Includes assessments of depression, anxiety, and composite measures of psychological wellbeing (eg, General Health Questionnaire). Where mental health subscales of quality-of-life measures were reported (eg, the mental composite score of the short form 36 (SF 36), these were included in this category. Includes both generic and disease specific measures of quality of life. Where composite scales such as SF 36 were reported as a total score and as subscores, both were reported. Includes assessments such as exercise frequency, diet, and adherence to preventative treatments (eg, use of inhaler). **Includes measures such as numbers of hospital visits, visits to a health-care professional, and use of medication for symptomatic relief (eg, pain relief). Table 4: Outcome measured in self-management studies for arthritis, asthma, and type 2 diabetes the apparent difference between these two forms of arthritis are unclear, but comparison is complicated because content of interventions tends to differ between the two illnesses, and studies in rheumatoid arthritis are generally smaller and more likely to be underpowered than those for osteoarthritis. Three studies with mixed osteoarthritis and rheumatoid arthritis groups assessed versions of the Arthritis Self-Management Programme. 119 No effect on pain or disability was evident in the UK 86 or US primary care 87 studies; however, the Spanish-language adaptation 85 reported benefits in both pain and disability. This version was not a direct translation of the original programme but included additional features such as practice of exercise in the classes. Of note, exercise might be a crucial component in osteoarthritis interventions, 120 so further exploration of its role seems to be important. Eight of the studies 88,92,93,95,96,103,106,108 for asthma measured severity of symptoms and frequency or the percentage of symptom-free days. The three studies that recorded reductions in severity all used education and action plans. 92,95,96 No study identified an increase in the percentage of symptom-free days. Assessment of psychological wellbeing as an outcome is complicated by selection of participants. Although numbers with depression and anxiety tend to be higher than in the general population, many individuals recruited into self-management programmes might show little evidence of depressed mood or increased anxiety. Expectations that these outcomes will improve after an SMI might therefore be unrealistic. 121 Of the ten studies for diabetes that measured these outcomes, seven reported no differences compared with controls; the two studies that measured this outcome for asthma likewise found no differences compared with controls. By contrast, for arthritis, six of ten studies that measured psychological wellbeing reported benefits. All of those that used cognitive behavioural programmes 70,76,77,80 and one of five based on social learning theory 86 recorded improvement in psychological wellbeing. A multidisciplinary training programme that included cognitive behavioural and social-learning components also reported benefits. 72 For diabetes, SMIs that showed benefits in psychological wellbeing also tended to include some cognitive behavioural components, 47,50 suggesting that programmes that are more cognitive are more likely to yield improvements in psychological wellbeing for both arthritis and diabetes. was assessed in 12 (67%) of studies for asthma, but despite its widespread use in health care it was assessed in only seven (33%) for diabetes and seven (29%) for arthritis. However, measurement of quality of life for arthritis is complicated by the fact that disease-specific instruments are widely used and their subscales are generally reported under symptoms, function, and psychological wellbeing rather than as a composite measure. For both diabetes and arthritis, little effect was recorded for quality of life. One study for diabetes that reported clear benefits 61 used a long intervention (4 years) and found that changes emerged only after the 24 month follow-up. This study highlights that some changes in outcomes might only develop after a long time. Half of the studies (six) for asthma that measured quality of life, generally using an asthma-specific measure, reported significant benefits. All were based on education and a form of action plan. Importantly, three studies that failed to find such improvements were also based on the same approach; therefore, no particular pattern is discernible. These findings suggest that in all three illnesses, the relation between SMIs and quality of life is not well understood. The changes in behaviour needed by SMIs might constrain quality of life, but the absence of evidence of such deterioration in our review suggests that SMIs have no real cost to patients quality of life. Further research is needed to understand the complex relation between SMIs and quality of life. Improvement of self-management behaviours, such as diet and exercise, or more cognitive behaviours, such as effective coping, is a prime focus of these types of interventions, but several did not assess behaviour for all three illnesses (diabetes eight [38%]; arthritis 12 [50%]; asthma four [22%]). These studies seem to have assumed that a simple relation exists between behaviour change and other outcomes, but it is extremely complex. 122 This fact is illustrated by studies in which behaviour changed in the absence of changes in more clinical and symptomatic measures, 58,67,68,86,88,106 and by cases in which changes in clinical or symptomatic measures happened in the absence of measurable changes in behaviour. 62,63,89,102 Some of these findings could be a result of the manner in which behaviour was assessed and the threshold of behaviour changes leading to clinical or symptomatic change as Vol 364 October 23, 2004

11 well as the multiple aspects of behaviour measured in some studies. Nonetheless, it is clear that there is no simple one-to-one relation between behaviour and measures of symptoms and clinical state. Assessment of behaviour would therefore seem valuable to further understand this relation. Seven of 13 diabetes studies that measured the effect of SMIs on self-management behaviours indicated some change compared with a control group, 43,44,49,52,58,61,111 and four further studies showed changes in behaviour over time. 55,57,63,111 Alterations in diet and exercise were the most frequently measured behaviours. The findings suggest that changes in behaviour do happen after SMIs for diabetes, and are similar to those reported in the review by Norris and others, 39 who found a positive effect of self-management training on self-care and lifestyle behaviours. Of 12 studies that assessed changes in behaviour for arthritis, ten reported some evidence of change. 65,67,68,72,74,75,83 86 Seven 65,67,68,74,84 86 showed increases in exercise, joint protection, or both, compared with controls, and two showed changes over time in these behaviours. 72,83 Most of these interventions were based on a social-learning approach. Four studies for arthritis 67,69,75,77 assessed adherence, but only two specifically targeted adherence to medication in the intervention, 67,75 of which one 75 reported a positive outcome. Many (14 [78%]) asthma interventions targeted some aspect of behaviour, usually adherence to preventive medications, recognition and appropriate use of rescue medications as well as inhaler technique, self-monitoring, and avoidance of asthma triggers. Significant changes in behaviour were reported in eight (57%) of these studies, six of which used an education and action-plan approach. change, the focus of SMIs, is the most successful outcome assessed in these studies. Importantly, the behaviours examined are very different in the three illnesses. Some, as in asthma, are specifically related to the illness, whereas others such as diet and exercise are more integrated with lifestyle and potentially more difficult to change. The fact that other variables did not change as a result of behaviour change in some studies makes it important to consider the relation between behaviour and more clinical outcomes, and raises the important question of whether changes in behaviour should be the key outcome for SMIs. Reduction in use of health care is one of the possible economic benefits of SMIs, and was most frequently assessed for asthma but not at all for diabetes (table 4). 64% (seven of 11) of interventions for asthma showed reductions in such use, and six of these used an education and action-plan approach. 92,94,95,98,100,103 Only one study that used this approach and assessed this outcome did not find any reduction. 106 This finding suggests that the education and action-plan approach is effective at reducing use of health care in asthma (see also Gibson and colleagues 115 ). By contrast with asthma, SMIs for arthritis and diabetes were less likely to have an immediate effect on use of health care, since control of symptoms to restrict emergency visits is not the focus of these interventions. Nonetheless, improved self-management could change use of health care. Six studies examined such use for arthritis, including all four that combined both rheumatoid arthritis and osteoarthritis. Two found some reduction in visits to health care professionals. 83,84 An enduring issue for SMIs is the duration of any effects observed. Studies on asthma tended to follow up patients for longer than those for both diabetes and arthritis, perhaps partly because asthma studies tended to measure use of health care, such as emergency room visits, and these tend to need a long period of assessment to obtain some range in results. Few SMIs for diabetes and arthritis assessed findings for more than 12 months, and in those that did, many showed that benefits tend not to be retained at long-term followup, although this might be changing for diabetes. 60,61 Although expectation of long-term effects from SMIs might be unreasonable because of the short-term nature of many of the interventions, it remains important not only to examine whether people are able to adopt self-management behaviours in the long term, but also to devise techniques that can lead to long-term change in behaviour. In this context, the possible benefits of booster sessions to reinforce such change need more examination. Discussion The SMIs for each of the three illnesses conditions in this review have general differences in content, which seem to result from the different objectives for management of each illness. Some of the studies for asthma attempted to halt development of symptoms, whereas those for arthritis tended to try to reduce the effect of the symptoms and illness, and therefore tended to have broader focus. For diabetes, many of the studies were concerned with aspects of integration of the complex regimen into patients lifestyles. As such, these SMIs also tended to be broader than those for asthma. One approach to definition of self management 5 could consider the studies of asthma as intensive management of symptoms rather than true self management. Alternatively, to the extent that these interventions engage patients in a further attempt to manage and take control over their illness to avoid a major effect on quality of life, they may be classed as self management as opposed to disease management. Whether or not patients with asthma could benefit further by broadening of the interventions to encompass emotional aspects of coping remains to be established. An alternative approach to the disease-specific SMIs considered in this review is to adopt a generic approach based on the premise that a core generic set of skills eg, problem solving and goal setting might be Vol 364 October 23,

12 sufficient to improve management. The chronic disease self-management programme 123,124 takes this approach and includes people with different chronic illnesses in the same self-management group. Establishing the extent to which these generic rather than diseasespecific skills are sufficient for the desired change in any illness would be an important step in the developing field of self management. Although there were general differences between illnesses in the content of SMIs, the range and diversity within illnesses was also large. This range raises the challenge of how to interpret the findings within each illness. Some reviews of SMIs have used metaanalysis, 117,125 but its applicability has to be questioned when interventions are so diverse, as illustrated in the tables. By incorporating all SMIs into one analysis, one cannot find out which of the different types of intervention are effective and for which outcomes. A more discursive approach that examines the nature and content of the studies that showed positive results might be a more pragmatic method of advancing our understanding in this specialty. When considering self management for diabetes, arthritis, and asthma, it is important to bear in mind the limitations of our review. We considered only studies published between 1997 and Although the studies were assessed against several outcomes, not all outcomes (in particular all clinical outcomes) were included. Furthermore, psychosocial constructs, which could explain how change happens in the intervention, were not considered because of space limitations. We have also not considered the relative sensitivity of the instruments used to assess change. However, the examination of the processes leading to change after SMIs remains an important area of study, not only to understand the processes at work, but also to examine differences between participants. This review is also likely to present a fairly optimistic view of SMIs, since most measured several outcomes, which increases the chance of a positive outcome. Moreover, where differences between groups were not available, changes over time were also considered. Changes over time are subject to attentional influences and are a much weaker design than those studies with a control group. A limiting factor when attempting to make sense of this area is the description of the interventions in reports. Previous reviews have criticised studies for not fully describing the content of the interventions and providing little detail of their theoretical backgrounds. 126,127 As has been suggested, 39,127 this problem could be overcome if authors published protocols and manuals of their programmes; this would not only allow accurate interpretation of the intervention but would also help ensure that within a programme, each facilitator delivers the same content in a consistent manner. Descriptions of standard care should also be offered, particularly if this consists of some form of education, as is commonly the case for diabetes. This description is important because of the likely variation between locations and the changes in standard practice over time as care of these illnesses evolves. Finally, our review was limited to randomised trials. Although randomised control trials are the gold standard for assessment of interventions in health care, there is debate as to whether this is the most suitable method for SMIs. One criticism of these designs is that the overriding focus is efficacy, with less attention to factors that could influence effectiveness 128 ie, whether the benefits of interventions are still evident when integrated into wider health care services (eg, Glasgow and others 129 ). One mechanism to assist integration is recognition of the need for SMIs within national standards of care. Increasingly, such standards do endorse the need for self management (see asthma in both the UK 130 and the USA, 131 and diabetes in the USA 132 ). For SMIs to have greater uptake, thought should be given to how and when they are offered to patients. Introduction and endorsement of these programmes at a physician visit will probably ensure higher rates of participation. It should also be recognised, however, that as with medication, one therapy or programme might not be suitable for all patients. Identifying who benefits most from which SMIs is an important addition to any assessment, and could lead to more effective targeting of resources. If SMIs are to be more widely adopted in health care, training in skills such as group facilitation, problem solving, goal setting, and cognitive-behavioural techniques need to be enhanced; they are not usually part of most health-care professionals training. If these interventions are to be delivered appropriately and effectively, training in the skills needed by health-care professionals who deliver these programmes needs to be recognised and appropriate courses developed. References 1 World Population Prospects: The 2000 Revision. Highlights. New York: United Nations, Population Division Department of Economic and Social Affairs, Holman H, Lorig K. Patients as partners in managing chronic disease. Partnership is a prerequisite for effective and efficient health care. BMJ 2000; 320: Murray CJL. The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to Los Angeles: Harvard University Press, Epping-Jordan J, Bengoa R, Kawar R, Sabate E. The challenge of chronic conditions: WHO responds. BMJ 2001; 323: Barlow J, Wright C, Sheasby J, Turner A, Hainsworth J. Selfmanagement approaches for people with chronic conditions: a review. Patient Educ Couns 2002; 48: Glasgow RE. Compliance to Diabetes Regimens: Conceptualiation, Complexity and Determinants. In: Cramer JA, Spilker B, eds. Patient compliance in medical practice. New York: Raven Press, 1991: Cline CM, Bjorck-Linne AK, Israelsson BY, Willenheimer RB, Erhardt LR. Non-compliance and knowledge of prescribed medication in elderly patients with heart failure. Eur J Heart Fail 1999; 1: Vol 364 October 23, 2004

13 8 Cochrane GM, Horne R, Chanez P. Compliance in asthma. Respir Med 1999; 93: Kravitz RL, Hays RD, Sherbourne CD, et al. Recall of recommendations and adherence to advice among patients with chronic medical conditions. Arch Intern Med 1993; 153: Peterman AH, Cella DF. Adherence issues among cancer patients. In: Schumaker SA, Schron EB, Ockene JK, eds. The handbook of health behaviour change. 2nd edn. New York: Springer, 1998: Rubin RR, Peyrot M. and diabetes. Diabetes Metab Res Rev 1999; 15: van t Spijker A, Trijsburg RW, Duivenvoorden HJ. Psychological sequelae of cancer diagnosis: a meta-analytical review of 58 studies after Psychosom Med 1997; 59: Pincus T, Griffith J, Pearce S, Isenberg D. Prevalence of selfreported depression in patients with rheumatoid arthritis. Br J Rheumatol 1996; 35: Hawley DJ, Wolfe F. Depression is not more common in rheumatoid arthritis: a 10-year longitudinal study of 6153 patients with rheumatic disease. J Rheumatol 1993; 20: Lehrer P, Feldman J, Giardino N, Song HS, Schmaling K. Psychological aspects of asthma. J Consult Clin Psychol 2002; 70: Juenger J, Schellberg D, Kraemer S, et al. Health related quality of life in patients with congestive heart failure: comparison with other chronic diseases and relation to functional variables. Heart 2002; 87: Parcel GS, Bartlett EE, Bruhn JG. The role of health education in self-management. In: Holroyd KA, Creer TL, eds. Self-management of chronic disease: handbook of clinical interventions and research. London: Academic Press, 1986: Gibson P, Powell H, Coughlan J, et al. Limited (information only) patient education programs for adults with asthma. Cochrane Database Syst Rev 2002: CD Taal E, Rasker JJ, Wiegman O. Group education for rheumatoid arthritis patients. Semin Arthritis Rheum 1997; 26: Coates VE, Boore JR. Knowledge and diabetes self-management. Patient Educ Couns 1996; 29: Bandura A. Social foundations of thought and action: a social cognitive theory. NJ, USA: Prentice-Hall, Lazarus RS. Coping with the stress of illness. WHO Reg Publ Eur Ser 1992; 44: Folkman S. Personal control and stress and coping processes: a theoretical analysis. J Pers Soc Psychol 1984; 46: Prochaska JO, Velicer WF. The transtheoretical model of health behavior change. Am J Health Promot 1997; 12: Prochaska JO, DiClemente CC, Velicer WF, Rossi JS. Criticisms and concerns of the transtheoretical model in light of recent research. Br J Addict 1992; 87: DeRubeis RJ, Crits-Christoph P. Empirically supported individual and group psychological treatments for adult mental disorders. J Consult Clin Psychol 1998; 66: Morley S, Eccleston C, Williams A. Systematic review and metaanalysis of randomized controlled trials of cognitive behaviour therapy and behaviour therapy for chronic pain in adults, excluding headache. Pain 1999; 80: Keefe FJ, Caldwell DS. Cognitive Behavioral Interventions. In: Wegener ST, Belza BL, Gall EP, eds. Clinical care in the rheumatic diseases. Atlanta: American College of Rheumatology, Denollet J, Brutsaert DL. Reducing emotional distress improves prognosis in coronary heart disease: 9-year mortality in a clinical trial of rehabilitation. Circulation 2001; 104: Snoek FJ, Skinner TC. Psychological counselling in problematic diabetes: does it help? Diabet Med 2002; 19: Bodenheimer T, Lorig K, Holman H, Grumbach K. Patient selfmanagement of chronic disease in primary care. JAMA 2002; 288: Von Korff M, Gruman J, Schaefer J, Curry SJ, Wagner EH. Collaborative management of chronic illness. Ann Intern Med 1997; 127: King H, Aubert RE, Herman WH. Global burden of diabetes, : prevalence, numerical estimates, and projections. Diabetes Care 1998; 21: Lawrence RC, Helmick CG, Arnett FC, et al. Estimates of the prevalence of arthritis and selected musculoskeletal disorders in the United States. Arthritis Rheum 1998; 41: World Health Organisation. Bronchial asthma. Fact Sheet No 206. Revised January Available at: (accessed 13 Sept, 2004) 36 Smith DH, Malone DC, Lawson KA, Okamoto LJ, Battista C, Saunders WB. A national estimate of the economic costs of asthma. Am J Respir Crit Care Med 1997; 156: American Diabetes Association. Economic consequences of diabetes mellitus in the US in Diabetes Care 1998; 21: Harish Z, Bregante AC, Morgan C, et al. A comprehensive innercity asthma program reduces hospital and emergency room utilization. Ann Allergy Asthma Immunol 2001; 86: Norris SL, Engelgau MM, Narayan KM. Effectiveness of selfmanagement training in type 2 diabetes: a systematic review of randomized controlled trials. Diabetes Care 2001; 24: Lorig K, Konkol L, Gonzalez V. Arthritis patient education: a review of the literature. Patient Educ Couns 1987; 10: Sudre P, Jacquemet S, Uldry C, Perneger TV. Objectives, methods and content of patient education programmes for adults with asthma: systematic review of studies published between 1979 and Thorax 1999; 54: Neinstein LS. The treatment of adolescents with a chronic illness: physicians must go beyond the medical model. West J Med 2001; 175: Agurs-Collins TD, Kumanyika SK, Have TRT, Adams-Campbell LL. A randomized controlled trial of weight reduction and exercise for diabetes management in older African- American subjects. Diabetes Care 1997; 20: Smith DE, Heckemeyer CM, Kratt PP, Mason DA. Motivational interviewing to improve adherence to a behavioral weight-control program for older obese women with NIDDM. A pilot study. Diabetes Care 1997; 20: Samaras K, Ashwell S, Mackintosh AM, Fleury AC, Campbell LV, Chisholm DJ. Will older sedentary people with non-insulindependent diabetes mellitus start exercising? A health promotion model. Diabetes Res Clin Pract 1997; 37: Jablon SL, Naliboff BD, Gilmore SL, Rosenthal MJ. Effects of relaxation training on glucose tolerance and diabetic control in type II diabetes. Appl Psychophysiol Biofeedback 1997; 22: Henry JL, Wilson PH, Bruce DG, Chisholm DG, Rawling PJ. Cognitive behavioural stress-management for patients with noninsulin dependent diabetes mellitus. Psychol Health Med 1997; 2: Aikens JE, Kiolbasa TA, Sobel R. Psychological predictors of glycemic change with relaxation training in non-insulin-dependent diabetes mellitus. Psychother Psychosom 1997; 66: Vazquez IM, Millen B, Bissett L, Levenson SM, Chipkin SR. Buena Alimentacion, Buena Salud: a preventive nutrition intervention in Caribbean Latinos with type 2 diabetes. Am J Health Promot. 1998; 13: Lustman PJ, Griffith LS, Freedland KE, Kissel SS, Clouse RE. Cognitive behavior therapy for depression in type 2 diabetes mellitus. A randomized, controlled trial. Ann Intern Med 1998; 129: Ridgeway NA, Harvill DR, Harvill LM, Falin TM, Forester GM, Gose OD. Improved control of type 2 diabetes mellitus: a practical education/behavior modification program in a primary care clinic. South Med J 1999; 92: Kirk AF, Higgins LA, Hughes AR, et al. A randomized, controlled trial to study the effect of exercise consultation on the promotion of physical activity in people with Type 2 diabetes: a pilot study. Diabet Med 2001; 18: McKay HG, King D, Eakin EG, Seeley JR, Glasgow RE. The diabetes network internet-based physical activity intervention. Diabetes Care 2001; 24: Miller CK, Edwards L, Kissling G, Sanville L. Nutrition education improves metabolic outcomes among older adults with diabetes mellitus: results from a randomized controlled trial. Prev Med 2002; 34: McKay HG, Glasgow RE, Feil EG, Boles SM, Barrera M. Internetbased diabetes self-management and support: initial outcomes Vol 364 October 23,

14 from the Diabetes Network Project. Rehabil Psychol 2002; 47: Levetan C, Dawn K, Robbins D, Ratner R. Impact of computergenerated personalized goals on HbA(1c). Diabetes Care 2002; 25: Kenardy J, Mensch M, Bowen K, Green B, Walton J. Group therapy for binge eating in Type 2 diabetes: a randomized trial. Diabet Med 2002; 19: Keyserling TC, Ammerman AS, Samuel-Hodge CD, et al. A diabetes management program for African American women with type 2 diabetes. Diabetes Educ 2000; 26: Glasgow RE, Toobert DJ. Brief, computer-assisted diabetes dietary self-management counseling: effects on behavior, physiologic outcomes, and quality of life. Med Care 2000; 38: Brown SA, Garcia AA, Kouzekanani K, Hanis CL. Culturally competent diabetes self-management education for Mexican Americans: the Starr County border health initiative. Diabetes Care 2002; 25: Trento M, Passera P, Bajardi M, et al. Lifestyle intervention by group care prevents deterioration of type II diabetes: a 4-year randomized controlled clinical trial. Diabetologia 2002; 45: Surwit RS, van Tilburg MA, Zucker N, et al. Stress management improves long-term glycemic control in type 2 diabetes. Diabetes Care 2002; 25: Rickheim PL, Weaver TW, Flader JL, Kendall DM. Assessment of group versus individual diabetes education: a randomized study. Diabetes Care 2002; 25: Multon KD, Parker JC, Smarr KL, et al. Effects of stress management on pain behavior in rheumatoid arthritis. Arthritis Rheum 2001; 45: Lindroth Y, Brattstrom M, Bellman I, et al. A problem-based education program for patients with rheumatoid arthritis: evaluation after three and twelve months. Arthritis Care Res 1997; 10: Bell MJ, Lineker SC, Wilkins AL, Goldsmith CH, Badley EM. A randomized controlled trial to evaluate the efficacy of community based physical therapy in the treatment of people with rheumatoid arthritis. J Rheumatol 1998; 25: Brus HL, van de Laar MA, Taal E, Rasker JJ, Wiegman O. Effects of patient education on compliance with basic treatment regimens and health in recent onset active rheumatoid arthritis. Ann Rheum Dis 1998; 57: Hammond A, Lincoln N, Sutcliffe L. A crossover trial evaluating an educational-behavioural joint protection programme for people with rheumatoid arthritis. Patient Educ Couns 1999; 37: Helliwell PS, O Hara M, Holdsworth J, Hesselden A, King T, Evans P. A 12-month randomized controlled trial of patient education on radiographic changes and quality of life in early rheumatoid arthritis. Rheumatology 1999; 38: Leibing E, Pfingsten M, Bartmann U, Rueger U, Schuessler G. Cognitive-behavioral treatment in unselected rheumatoid arthritis outpatients. Clin J Pain 1999; 15: Lundgren S, Stenstrom CH. Muscle relaxation training and quality of life in rheumatoid arthritis. A randomized controlled clinical trial. Scand J Rheumatol 1999; 28: Scholten C, Brodowicz T, Graninger W, et al. Persistent functional and social benefit 5 years after a multidisciplinary arthritis training program. Arch Phys Med Rehabil 1999; 80: Smyth JM, Stone AA, Hurewitz A, Kaell A. Effects of writing about stressful experiences on symptom reduction in patients with asthma or rheumatoid arthritis: a randomized trial. J Am Med Assoc 1999; 281: Hammond A, Freeman K. One-year outcomes of a randomized controlled trial of an educational-behavioural joint protection programme for people with rheumatoid arthritis. Rheumatology 2001; 40: Hill J, Bird H, Johnson S. Effect of patient education on adherence to drug treatment for rheumatoid arthritis: a randomised controlled trial. Ann Rheum Dis 2001; 60: Sharpe L, Sensky T, Timberlake N, Ryan B, Brewin CR, Allard S. A blind, randomized, controlled trial of cognitive-behavioural intervention for patients with recent onset rheumatoid arthritis: preventing psychological and physical morbidity. Pain 2001; 89: Evers AW, Kraaimaat FW, van Riel PL, de Jong AJ. Tailored cognitive-behavioral therapy in early rheumatoid arthritis for patients at risk: a randomized controlled trial. Pain 2002; 100: Freeman K, Hammond A, Lincoln NB. Use of cognitivebehavioural arthritis education programmes in newly diagnosed rheumatoid arthritis. Clin Rehabil 2002; 16: Cronan TA, Groessl E, Kaplan RM. The effects of social support and education interventions on health care costs. Arthritis Care Res 1997; 10: Keefe FJ, Caldwell DS, Baucom D,. Spouse-assisted coping skills training in the management of knee pain in osteoarthritis: longterm follow-up results. Arthritis Care Res 1999; 12: Lord J, Victor C, Littlejohns P, Ross FM, Axford JS. Economic evaluation of a primary care-based education programme for patients with osteoarthritis of the knee. Health Technol Assess 1999; 3: Maurer BT, Stern AG, Kinossian B, Cook KD, Schumacher HR. Osteoarthritis of the knee: isokinetic quadriceps exercise versus an educational intervention. Arch Phys Med Rehabil 1999; 80: Hopman-Rock M, Westhoff MH. The effects of a health educational and exercise program for older adults with osteoarthritis for the hip or knee. J Rheumatol 2000; 27: Fries JF, Carey C, McShane DJ. Patient education in arthritis: randomized controlled trial of a mail-delivered program. J Rheumatol 1997; 24: Lorig K, Gonzalez VM, Ritter P. Community-based Spanish language arthritis education program: a randomized trial. Med Care 1999; 37: Barlow JH, Turner AP, Wright CC. A randomized controlled study of the arthritis self-management programme in the UK. Health Educ Res 2000; 15: Solomon DH, Warsi A, Brown-Stevenson T, et al. Does selfmanagement education benefit all populations with arthritis? A randomized controlled trial in a primary care physician network. J Rheumatol 2002; 29: Berg J, Dunbar Jacob J, Sereika S. An evaluation of a selfmanagement program for adults with asthma. Clin Nurs Res 1997; 6: Kauppinen R, Sintonen H, Tukiainen H. One-year economic evaluation of intensive vs conventional patient education and supervision for self-management of new asthmatic patients. Respir Med 1998; 92: Kauppinen R, Sintonen H, Vilkka V, Tukiainen H. Long-term (3- year) economic evaluation of intensive patient education for selfmanagement during the first year in new asthmatics. Respir Med 1999; 93: Kauppinen R, Vilkka V, Sintonen H, Klaukka T, Tukiainen H. Long-term economic evaluation of intensive patient education during the first treatment year in newly diagnosed adult asthma. Respir Med 2001; 95: Turner MO, Taylor D, Bennett R, Fitzgerald IM. A randomized trial comparing peak expiratory flow and symptom selfmanagement plans for patients with asthma attending a primary care clinic. Am J Respir Crit Care Med 1998; 157: Bailey WC, Kohler CL, Richards JM, et al. Asthma selfmanagement: do patient education programs always have an impact? Arch Intern Med 1999; 159: George MR, O Dowd LC, Martin I, et al. A comprehensive educational program improves clinical outcome measures in innercity patients with asthma. Arch Intern Med 1999; 159: de Oliveira MA, Faresin SM, Bruno VF, de Bittencourt AR, Fernandes AL. Evaluation of an educational programme for socially deprived asthma patients. Eur Respir J 1999; 14: Cote J, Cartier A, Robichaud PH, et al. Influence of asthma education on asthma severity, quality of life and environmental control. Can Respir J 2000; 7: Gallefoss F, Bakke PS, Kjaersgaard P. assessment after patient education in a randomized controlled study on asthma and chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1999; 159: Gallefoss F, Bakke PS. Impact of patient education and selfmanagement on morbidity in asthmatics and patients with chronic obstructive pulmonary disease. Respir Med 2000; 94: Vol 364 October 23, 2004

15 99 Gallefoss F, Bakke PS. Cost-effectiveness of self-management in asthmatics: a 1-yr follow-up randomized, controlled trial. Eur Respir J 2001; 17: Levy ML, Robb M, Allen J, Doherty C, Bland JM, Winter RJD. A randomized controlled evaluation of specialist nurse education following accident and emergency department attendance for acute asthma. Respir Med 2000; 94: Blixen CE, Hammel JP, Murphy D, Ault V. Feasibility of a nurserun asthma education program for urban African-Americans: a pilot study. J Asthma 2001; 38: Cote J, Bowie DM, Robichaud P, Parent J, Battisti L, Boulet L. Evaluation of two different educational interventions for adult patients consulting with an acute asthma exacerbation. Am J Respir Crit Care Med 2001; 163: Klein JJ, van der Palen J, Uil SM, Zielhuis GA, Seydel ER, van Herwaarden CLA. Benefit from the inclusion of self-treatment guidelines to a self-management programme for adults with asthma. Eur Respir J 2001; 17: van der Palen J, Klein JJ, Zielhuis GA, van Herwaarden CLA, Seydel ER. al effect of self-treatment guidelines in a selfmanagement program for adults with asthma. Patient Educ Couns 2001; 43: Schmaling KB, Blume AW, Afari N. A randomized controlled pilot study of motivational interviewing to change attitudes about adherence to medications for asthma. J Clin Psychol Med Settings 2001; 8: Couturaud F, Proust A, Frachon I, et al. Education and selfmanagement: a one-year randomized trial in stable adult asthmatic patients. J Asthma 2002; 39: Hockemeyer J, Smyth J. Evaluating the feasibility and efficacy of a self-administered manual-based stress management intervention for individuals with asthma: results from a controlled study. Behav Med 2002; 27: Marabini A, Brugnami G, Curradi F, et al. Short-term effectiveness of an asthma educational program: results of a randomized controlled trial. Respir Med 2002; 96: Perneger TV, Sudre P, Muntner P, et al. Effect of patient education on self-management skills and health status in patients with asthma: a randomized trial. Am J Med 2002; 113: Glasgow RE, Osteen VL. Evaluating diabetes education. Are we measuring the most important outcomes? Diabetes Care 1992; 15: Glasgow RE, Toobert DJ, Hampson SE, Strycker LA. Implementation, generalization and long-term results of the choosing well diabetes self-management intervention. Patient Educ Couns 2002; 48: Lorig K, Feigenbaum P, Regan C, Ung E, Chastain RL, Holman HR. A comparison of lay-taught and professional-taught arthritis self-management courses. J Rheumatol 1986; 13: Cohen JL, Sauter SV, devellis RF, devellis BM. Evaluation of arthritis self-management courses led by laypersons and by professionals. Arthritis Rheum 1986; 29: Griffin S, Kinmouth AL, Skinner C, Kelly J. Educational and psychosocial intervention for adults with diabetes: report to the British Diabetic Association. London: British Diabetic Association, Gibson PG, Powell H, Coughlan J, et al. Self-management education and regular practitioner review for adults with asthma. Cochrane Database Syst Rev 2003; (1): CD Hawley DJ. Psycho-educational interventions in the treatment of arthritis. Bailliere s Clinical Rheumatology 1995; 9: Riemsma RP, Kirwan JR, Taal E, Rasker JJ. Patient education for adults with rheumatoid arthritis. Oxford: Update Software, Astin JA, Beckner W, Soeken K, Hochberg MC, Berman B. Psychological interventions for rheumatoid arthritis: a meta-analysis of randomized controlled trials. Arthritis Rheum 2002; 47: Lorig K. Arthritis self-help course. HMO Pract 1995; 9: Thomas KS, Muir KR, Doherty M, Jones AC, O Reilly SC, Bassey EJ. Home based exercise programme for knee pain and knee osteoarthritis: randomised controlled trial. BMJ 2002; 325: Steed L, Lankester J, Jenkins C, et al. Evaluations of a selfmanagement programme for patients with type 2 diabetes and microalbuminuria. Diabet Med 2002; 19: Lorig K, Seleznick M, Lubeck D, Ung E, Chastain RL, Holman HR. The beneficial outcomes of the arthritis self-management course are not adequately explained by behavior change. Arthritis Rheum 1989; 32: Lorig KR, Sobel DS, Ritter PL, Laurent D, Hobbs M. Effect of a selfmanagement program on patients with chronic disease. Eff Clin Pract 2001; 4: Lorig KR, Sobel DS, Stewart AL, et al. Evidence suggesting that a chronic disease self-management program can improve health status while reducing hospitalization: a randomized trial. Med Care 1999; 37: Brown SA. Meta-analysis of diabetes patient education research: variations in intervention effects across studies. Res Nursing Health 1992; 15: Padgett D, Mumford E, Hynes M, Carter R. Meta-analysis of the effects of educational and psychosocial interventions on management of diabetes mellitus. J Clin Epidemiol 1988; 41: Brown SA. Interventions to promote diabetes self-management: state of the science. Diabetes Educ 1999; 25: Norris SL, Lau J, Smith SJ, Schmid CH, Engelgau MM. Self-management education for adults with type 2 diabetes: a meta-analysis of the effect on glycemic control. Diabetes Care 2002; 25: Glasgow RE, Wagner EH, Kaplan RM, Vinicor F, Smith L, Norman J. If diabetes is a public health problem, why not treat it as one? A population-based approach to chronic illness. Ann Behav Med 1999; 21: British guideline on the management of asthma. Thorax 2003; 58: Guidelines for the diagnosis and management of asthma. National Heart, Lung, and Blood Institute. National Asthma Education Program. Expert Panel Report. J Allergy Clin Immunol 1991; 88: Mensing C, Boucher J, Cypress M, et al. National standards for diabetes self-management education. Diabetes Care 2003; 26: Vol 364 October 23,